Environment for
DeveLoping
KDD-Applications
Supported by Index-Structures

de.lmu.ifi.dbs.elki.data.model
Class CorrelationAnalysisSolution<V extends RealVector<V,?>>

java.lang.Object
  extended by de.lmu.ifi.dbs.elki.data.model.CorrelationAnalysisSolution<V>
Type Parameters:
V - the type of RealVector handled by this Result
All Implemented Interfaces:
Model, Result, TextWriteable

public class CorrelationAnalysisSolution<V extends RealVector<V,?>>
extends Object
implements TextWriteable, Result, Model

A solution of correlation analysis is a matrix of equations describing the dependencies.

Author:
Arthur Zimek

Field Summary
private  Vector centroid
          The centroid if the objects belonging to the hyperplane induced by the correlation.
private  int correlationDimensionality
          The dimensionality of the correlation.
private  LinearEquationSystem linearEquationSystem
          Stores the solution equations.
private  NumberFormat nf
          Number format for output accuracy.
private  Matrix similarityMatrix
          The similarity matrix of the pca.
private  double standardDeviation
          The standard deviation within this solution.
private  Matrix strongEigenvectors
          The strong eigenvectors of the hyperplane induced by the correlation.
private  Matrix weakEigenvectors
          The weak eigenvectors of the hyperplane induced by the correlation.
 
Constructor Summary
CorrelationAnalysisSolution(LinearEquationSystem solution, Database<V> db, Matrix strongEigenvectors, Matrix weakEigenvectors, Matrix similarityMatrix, Vector centroid)
          Provides a new CorrelationAnalysisSolution holding the specified matrix.
CorrelationAnalysisSolution(LinearEquationSystem solution, Database<V> db, Matrix strongEigenvectors, Matrix weakEigenvectors, Matrix similarityMatrix, Vector centroid, NumberFormat nf)
          Provides a new CorrelationAnalysisSolution holding the specified matrix and number format.
 
Method Summary
 Matrix dataProjections(V p)
          Returns the data vectors after projection.
 Vector dataVector(V p)
          Returns the data vectors after projection.
 Matrix dataVectors(Matrix p)
          Returns the data vectors after projection.
private  double distance(Matrix p)
          Returns the distance of Matrix p from the hyperplane underlying this solution.
 double distance(V p)
          Returns the distance of RealVector p from the hyperplane underlying this solution.
 Vector errorVector(V p)
          Returns the error vectors after projection.
 Matrix errorVectors(Matrix p)
          Returns the error vectors after projection.
 Matrix errorVectors(V p)
          Returns the error vectors after projection.
 Vector getCentroid()
          Returns the centroid of this model.
 int getCorrelationDimensionality()
          Return the correlation dimensionality.
 LinearEquationSystem getNormalizedLinearEquationSystem(Normalization<V> normalization)
          Returns the linear equation system for printing purposes.
 Matrix getSimilarityMatrix()
          Returns the similarity matrix of the pca.
 double getStandardDeviation()
          Returns the standard deviation of the distances of the objects belonging to the hyperplane underlying this solution.
 Matrix getStrongEigenvectors()
          Returns a copy of the strong eigenvectors.
 Matrix getWeakEigenvectors()
          Returns a copy of the weak eigenvectors.
 void output(PrintStream outStream, Normalization<V> normalization, List<AttributeSettings> settings)
          Writes the clustering result to the given stream.
 void writeToText(TextWriterStream out, String label)
          Text output of the equation system
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

linearEquationSystem

private LinearEquationSystem linearEquationSystem
Stores the solution equations.


nf

private NumberFormat nf
Number format for output accuracy.


correlationDimensionality

private int correlationDimensionality
The dimensionality of the correlation.


standardDeviation

private final double standardDeviation
The standard deviation within this solution.


weakEigenvectors

private final Matrix weakEigenvectors
The weak eigenvectors of the hyperplane induced by the correlation.


strongEigenvectors

private final Matrix strongEigenvectors
The strong eigenvectors of the hyperplane induced by the correlation.


similarityMatrix

private final Matrix similarityMatrix
The similarity matrix of the pca.


centroid

private final Vector centroid
The centroid if the objects belonging to the hyperplane induced by the correlation.

Constructor Detail

CorrelationAnalysisSolution

public CorrelationAnalysisSolution(LinearEquationSystem solution,
                                   Database<V> db,
                                   Matrix strongEigenvectors,
                                   Matrix weakEigenvectors,
                                   Matrix similarityMatrix,
                                   Vector centroid)
Provides a new CorrelationAnalysisSolution holding the specified matrix.

Parameters:
solution - the linear equation system describing the solution equations
db - the database containing the objects
strongEigenvectors - the strong eigenvectors of the hyperplane induced by the correlation
weakEigenvectors - the weak eigenvectors of the hyperplane induced by the correlation
similarityMatrix - the similarity matrix of the underlying distance computations
centroid - the centroid if the objects belonging to the hyperplane induced by the correlation

CorrelationAnalysisSolution

public CorrelationAnalysisSolution(LinearEquationSystem solution,
                                   Database<V> db,
                                   Matrix strongEigenvectors,
                                   Matrix weakEigenvectors,
                                   Matrix similarityMatrix,
                                   Vector centroid,
                                   NumberFormat nf)
Provides a new CorrelationAnalysisSolution holding the specified matrix and number format.

Parameters:
solution - the linear equation system describing the solution equations
db - the database containing the objects
strongEigenvectors - the strong eigenvectors of the hyperplane induced by the correlation
weakEigenvectors - the weak eigenvectors of the hyperplane induced by the correlation
similarityMatrix - the similarity matrix of the underlying distance computations
centroid - the centroid if the objects belonging to the hyperplane induced by the correlation
nf - the number format for output accuracy
Method Detail

output

public void output(PrintStream outStream,
                   Normalization<V> normalization,
                   List<AttributeSettings> settings)
            throws UnableToComplyException,
                   IOException
Writes the clustering result to the given stream.

Parameters:
outStream - the stream to write to
normalization - Normalization to restore original values according to, if this action is supported - may remain null.
settings - the settings to be written into the header
Throws:
UnableToComplyException - if any feature vector is not compatible with values initialized during normalization
IOException

getNormalizedLinearEquationSystem

public LinearEquationSystem getNormalizedLinearEquationSystem(Normalization<V> normalization)
                                                       throws NonNumericFeaturesException
Returns the linear equation system for printing purposes. If normalization is null the linear equation system is returned, otherwise the linear equation system will be transformed according to the normalization.

Parameters:
normalization - the normalization, can be null
Returns:
the linear equation system for printing purposes
Throws:
NonNumericFeaturesException - if the linear equation system is not compatible with values initialized during normalization

getCorrelationDimensionality

public int getCorrelationDimensionality()
Return the correlation dimensionality.

Returns:
the correlation dimensionality

distance

public double distance(V p)
Returns the distance of RealVector p from the hyperplane underlying this solution.

Parameters:
p - a vector in the space underlying this solution
Returns:
the distance of p from the hyperplane underlying this solution

distance

private double distance(Matrix p)
Returns the distance of Matrix p from the hyperplane underlying this solution.

Parameters:
p - a vector in the space underlying this solution
Returns:
the distance of p from the hyperplane underlying this solution

errorVectors

public Matrix errorVectors(V p)
Returns the error vectors after projection.

Parameters:
p - a vector in the space underlying this solution
Returns:
the error vectors

errorVectors

public Matrix errorVectors(Matrix p)
Returns the error vectors after projection.

Parameters:
p - a vector in the space underlying this solution
Returns:
the error vectors

errorVector

public Vector errorVector(V p)
Returns the error vectors after projection.

Parameters:
p - a vector in the space underlying this solution
Returns:
the error vectors

dataProjections

public Matrix dataProjections(V p)
Returns the data vectors after projection.

Parameters:
p - a vector in the space underlying this solution
Returns:
the data projections

dataVectors

public Matrix dataVectors(Matrix p)
Returns the data vectors after projection.

Parameters:
p - a vector in the space underlying this solution
Returns:
the error vectors

dataVector

public Vector dataVector(V p)
Returns the data vectors after projection.

Parameters:
p - a vector in the space underlying this solution
Returns:
the error vectors

getStandardDeviation

public double getStandardDeviation()
Returns the standard deviation of the distances of the objects belonging to the hyperplane underlying this solution.

Returns:
the standard deviation of this solution

getStrongEigenvectors

public Matrix getStrongEigenvectors()
Returns a copy of the strong eigenvectors.

Returns:
a copy of the strong eigenvectors

getWeakEigenvectors

public Matrix getWeakEigenvectors()
Returns a copy of the weak eigenvectors.

Returns:
a copy of the weak eigenvectors

getSimilarityMatrix

public Matrix getSimilarityMatrix()
Returns the similarity matrix of the pca.

Returns:
the similarity matrix of the pca

getCentroid

public Vector getCentroid()
Returns the centroid of this model.

Returns:
the centroid of this model

writeToText

public void writeToText(TextWriterStream out,
                        String label)
Text output of the equation system

Specified by:
writeToText in interface TextWriteable
Parameters:
out - Output writer
label - Label

Release 0.2 (2009-07-06_1820)